Received: 7 March 2022 | Revised: 8 June 2022 | Accepted: 15 July 2022 DOI: 10.1111/jopy.12757 O R I G I N A L A R T I C L E A developmental perspective on personality– relationship transactions: Evidence from three nationally representative samples Janina Larissa Bühler1 | Marcus Mund2 | Franz J. Neyer3 | Cornelia Wrzus4 1Department of Psychology, Johannes Gutenberg University Mainz, Mainz, Abstract Germany Objective: Throughout their lives, people experience different relationship 2Institute of Psychology, University of events, such as beginning or dissolving a romantic relationship. Personality traits Klagenfurt, Klagenfurt, Austria predict the occurrence of such relationship events (i.e., selection effects), and re- 3Institute of Psychology, Friedrich Schiller University Jena, Jena, Germany lationship events predict changes in personality traits (i.e., socialization effects), 4Institute of Psychology, Ruprecht Karls summarized as personality– relationship transactions. So far, evidence was partly University of Heidelberg, Heidelberg, inconsistent as to how personality traits and relationship events are linked with Germany each other. In this article, we argue that unnoticed age differences might have led Correspondence to these inconsistencies. To systematically test for age differences in transactions, Janina Larissa Bühler, Department we conceptualize relationship events in terms of gains and losses and apply a of Psychology, Johannes Gutenberg University Mainz, Binger Str. 14- 16, developmental perspective on transactions. 55122 Mainz, Germany. Methods: Using longitudinal data from three nationally representative sam- Email: jbuehler@uni-mainz.de ples (SOEP, HILDA, Understanding Society), we computed event-f ocused latent Funding information growth models and summarized the results meta-a nalytically. Swiss National Science Foundation; Results: The findings indicated some transactions. Of these, selection effects JOHANNES GUTENBERG were stronger than socialization effects, and effects of gain-b ased events were UNIVERSITAET MAINZ stronger than effects of loss- based events. We observed few interactions with age. Conclusion: Selection effects and, particularly, socialization effects, tend to be rare and fairly independent of age. We discuss a series of broader and narrower factors that may have an impact on the strength of transactions across adulthood. K E Y W O R D S life- span development, personality development, personality–r elationship transactions, relationship events, romantic relationships 1 | INTRODUCTION that relationship events are linked with people's person- ality traits, that is, with their relatively enduring patterns Throughout their lives, people experience different of thoughts, feelings, strivings, and behaviors (John & events in their romantic relationship lives, such as be- Srivastava, 1999). Specifically, personality traits have been ginning or dissolving a relationship. Research suggests found to predict the occurrence of relationship events (i.e., This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2022 The Authors. Journal of Personality published by Wiley Periodicals LLC. Journal of Personality. 2022;00:1–20. wileyonlinelibrary.com/journal/jopy | 1 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2 | BÜHLER et al. selection effects), and relationship events have been found (see theory on person–e nvironment transactions, Caspi to predict changes in personality traits (i.e., socialization ef- et al., 2005; Roberts et al., 2008): People may actively se- fects). This interplay has been summarized as personality– lect themselves into an event, may be selected based on relationship transactions (e.g., Magnusson, 1990; Neyer & their personality into a particular social role by others, or Asendorpf, 2001). may leave environmental conditions that no longer match Although transactions occur across entire adulthood, their traits. For instance, evidence suggests that less agree- the specific interplay between personality traits and re- able people, compared to more agreeable people, get more lationship events might differ across age. For example, likely divorced (Solomon & Jackson, 2014). This view can patterns of selection and socialization effects may be dif- account for selection effects, but it is limited in explaining ferent for people entering their first romantic relationship socialization effects. in young adulthood compared to others who start their Environmental views, on the other hand, consider first relationship as middle- aged adults. Such differences personality as a malleable system that is open to be may be due to the age- graded normativeness associated shaped by environmental factors, such as by life events with the relationship event. Thus, in the present study we (see neo-s ocioanalytic theory, Roberts & Wood,  2006). propose that research on transactions would benefit from Evidence suggests that personality traits develop through considering the age- graded normativeness of relationship the normative commitments, social roles, and behavioral events, which, in turn, should matter for the strength of scripts that are associated with life events (Lodi-S mith selection and socialization effects (Neyer et al.,  2014). & Roberts, 2007; Roberts et al., 2008). For instance, re- Using data from three nationally representative samples, search provides evidence that divorced people become we test our hypotheses in each data set separately and somewhat less conscientious over time (Roberts & then meta- analytically summarize the study-l evel results, Bogg, 2004). This view more likely accounts for social- allowing for internal replications and robustness checks ization effects. (e.g., Duncan et al., 2014). At the same time, for any selection or socialization effects to occur, individuals and their environments have to interact with each other. For instance, if someone 1.1 | Transactions between personality wants to select a romantic partner, there must be po- traits and relationship events tential mates in the environment who are available and interested (Günaydin et al., 2013). Similarly, the relation- Life events, in general, are defined as “time-d iscrete tran- ship event of separation rarely is a stochastic- contextual sitions that mark the beginning or the end of a specific experience, but it is influenced by the characteristics of status” (Luhmann et al.,  2012, p. 594). Accordingly, re- the person, including their personality traits (Dyrenforth lationship events can be characterized as time- discrete et al., 2010). Hence, in the real (romantic) world, the cat- transitions that indicate the start or ending of a particu- egories of endogenous and environmental views are less lar status in the romantic relationship domain (see also dichotomous and less exclusive given that individuals Bleidorn et al., 2018). Relationship events are qualitatively and their environments depend on each other. Thus, to different from each other and can be classified into the fully understand the conditions that precede and result two broad domains of gain-b ased and loss-b ased events from personality traits, it is important to test selection (Denissen et al.,  2019): Gain- based relationship events and socialization effects simultaneously. Otherwise, se- imply that a particular relationship status has begun, in- lection effects might mask socialization effects, and vice cluding “beginning a romantic relationship”, “moving versa (e.g., Mund et al., 2018). in with a partner”, and “marriage”.1 Loss- based relation- ship events imply that a particular relationship status has ended, including “separation”, “divorce”, and “death of 1.2 | Selection effects and socialization the partner/widowhood”. As noted above, relationship effects: A brief review events and personality traits are transactionally linked through selection and socialization effects. Two promi- In this section, we briefly review the existing evidence nent perspectives discuss the mechanisms that may drive on selection and socialization effects between personal- these effects: endogenous and environmental views. ity traits and relationship events, differentiating between Endogenous views posit that personality traits predis- gain-b ased and loss-b ased events (Denissen et al., 2019). pose people to experience a certain event or not (see five- An overview of the evidence is given in Table S1, including factor theory of personality, McCrae & Costa Jr.,  2008). participants' mean age and methodological information There are various mechanisms by which personality on the study design (i.e., measure and time lag between traits may select individuals to experiencing an event assessments).2 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BÜHLER et al. | 3 1.2.1 | Selection effects with research on relationship development, indicating that lower agreeableness and higher neuroticism Gain- based and loss-b ased relationship events are among those traits to predict lower relationship Beginning a romantic relationship. Divergent findings satisfaction (Dyrenforth et al., 2010). Lower relationship exist for selection effects on beginning the first serious satisfaction, in turn, may eventually lead to separation romantic relationship. Regarding neuroticism or or, if married, to divorce (Karney & Bradbury,  1995; neuroticism-r elated aspects (i.e., depression), one study Kelly & Conley, 1987). observed that higher neuroticism predicted that people enter their first romantic relationship in the following Divorce. Whereas one study indicated that personality 8 years (e.g., Neyer & Lehnart,  2007), whereas another traits do not predict divorce (Lehnart & Neyer,  2006), study indicated that lower depression predicted that four other studies indicated that higher neuroticism people enter a relationship in the following 2 years (Solomon & Jackson,  2014), lower agreeableness (Wagner et al.,  2015). More consistency exists for (Asselmann & Specht, 2020; Solomon & Jackson, 2014), extraversion, showing that higher extraversion and lower conscientiousness (Roberts & Bogg, 2004; Solomon higher sociability predicted that people (Neyer & & Jackson,  2014), and higher openness (Denissen et Lehnart, 2007; Wagner et al., 2015), particularly emerging al., 2019; Solomon & Jackson, 2014) predicted divorce. adults (Pusch et al.,  2019), enter their first romantic relationship. In addition, higher conscientiousness Widowhood. Two studies showed that none of the Big predicted that people enter a relationship (Pusch Five traits predicted widowhood (Denissen et al.,  2019; et al.,  2019), whereas these studies did not indicate Specht et al., 2011), suggesting that there is no evidence selection effects of agreeableness and openness. for selection effects of personality traits on experiencing widowhood. Moving in with a partner. Higher neuroticism and lower agreeableness (Asselmann & Specht, 2020), higher extraversion (Asselmann & Specht,  2020; Pusch et 1.2.2 | Socialization effects al., 2019; Specht et al., 2011) and higher conscientiousness (Pusch et al.,  2019) predicted that people move in with Gain- based and loss- based relationship events their partner. Research did not suggest selection effects of Beginning a romantic relationship. Beginning a romantic openness. relationship, particularly the first serious romantic relationship, has consistently been found to relate to Marriage. One study indicated that lower neuroticism a decrease in neuroticism (Lehnart et al.,  2010; Neyer predicted that people marry (Denissen et al.,  2019), & Asendorpf,  2001;Neyer & Lehnart,  2007; Wagner while another study indicated that higher neuroticism et al.,  2015). Somewhat less consistently, starting a (among women) predicted that people marry (Specht et relationship has also been found to relate to an increase al.,  2011). In addition, lower agreeableness (Asselmann in extraversion (Neyer & Lehnart,  2007; Wagner et & Specht,  2020), higher conscientiousness (Denissen al.,  2015), an increase in conscientiousness (Neyer & et al.,  2019), and lower openness (Denissen et al.,  2019) Asendorpf, 2001; Wagner et al., 2015), and a decrease in predicted marriage. openness (Pusch et al.,  2019). Research did not suggest socialization effects on agreeableness. Separation. Whereas four studies did not observe personality traits to predict separation (Lehnart & Moving in with a partner. One study observed no Neyer,  2006; Neyer & Asendorpf,  2001; Neyer & socialization effects of moving in with a partner (Specht Lehnart,  2007; Specht et al.,  2011), three other studies et al., 2011), while three other studies observed significant reported selection effects on separation: Higher effects: Moving in with a partner predicted an increase neuroticism (Asselmann & Specht,  2020; Solomon & in agreeableness (Pusch et al.,  2019), an increase in Jackson, 2014), higher extraversion (Pusch et al., 2019), conscientiousness (Asselmann & Specht,  2020), and lower agreeableness (Asselmann & Specht,  2020; a decrease in openness (Pusch et al.,  2019). Research Pusch et al.,  2019; Solomon & Jackson,  2014), lower did not suggest socialization effects on neuroticism and conscientiousness (Solomon & Jackson,  2014), extraversion. and higher openness (Solomon & Jackson,  2014) predicted that individuals separate. The finding that Marriage. Two studies observed no associations between lower agreeableness and higher neuroticism yielded getting married and changes in personality traits (Denissen prospective effects on separation also corresponds et al.,  2019; Neyer & Asendorpf,  2001). In three other 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 4 | BÜHLER et al. studies, however, marriage revealed prospective effects samples from young and middle adulthood or did not on personality change: Marriage predicted decreases in specifically consider age-s pecific developmental patterns neuroticism (Costa Jr. et al., 2000), extraversion (Costa Jr. (although some studies considered age as covariate, see et al., 2000; Specht et al., 2011), and openness (Asselmann Asselmann & Specht, 2020; Denissen et al., 2019; Specht & Specht, 2020; Costa Jr. et al., 2000; Specht et al., 2011). et al., 2011). Hence, the observed inconsistencies in trans- actions may be partly driven by age effects. Separation. Whereas four studies indicated no socialization effects of separation on personality traits (Asselmann & Specht, 2020; Denissen et al., 2019; Neyer & 1.3 | Age matters: Transactions Asendorpf, 2001; Pusch et al., 2019), two studies observed across the life span prospective effects: Neyer and Lehnart (2007) found that separation predicted an increase in extraversion, while The reviewed evidence informs about personality– Specht et al.  (2011) found that separation predicted relationship transactions, but existing findings may be increases in agreeableness and openness (only in men). limited in their generalizability across adulthood due to the shortage of studies testing age- specific hypotheses Divorce. Whereas one study observed no socialization in age-h eterogeneous samples. For example, regarding effects of divorce on traits (Denissen et al.,  2019), five socialization effects, Neyer and Asendorpf  (2001) stated studies revealed prospective effects, but for different traits that “engaging in a [first] serious partnership is a game and in different directions: Specifically, divorce has been one can only win” (p. 1200). This conclusion, however, found to predict an increase in neuroticism (Asselmann & is drawn from a young-a dult sample (M  =  28.6 years, Specht, 2020), an increase (Costa Jr. et al., 2000) or decrease SD = 3.8 years) and, as the authors themselves stated, it (Allemand et al.,  2015) in extraversion, an increase in remains open whether engaging in the first romantic rela- agreeableness (Spikic et al.,  2021), a decrease (Costa Jr. tionship is a game that people of either age can only win, et al., 2000; Roberts & Bogg, 2004; Spikic et al., 2021) or or whether relationship events have different predictors increase (Specht et al.,  2011) in conscientiousness, and and different implications at different ages. Moreover, an increase in openness (Costa Jr. et al.,  2000; Spikic et even in samples of young adults, age-d ifferential socializa- al., 2021). tion effects may occur: Wagner et al. (2015) observed that engaging in the first romantic relationship was related to Widowhood. Whereas widowhood has not been found later personality (i.e., lower neuroticism, extraversion, to predict personality change in a representative Dutch conscientiousness, and self-e steem) between ages 23 and sample (Denissen et al.,  2019), gender- differential 25, but not between ages 21 and 23. Hence, to allow con- socialization effects emerged in a representative German clusions about the generalizability of transactions across sample (Specht et al.,  2011): Women decreased in adulthood, it is essential to account for age- differential ef- conscientiousness after having lost their spouse, whereas fects, both theoretically and empirically. men increased in conscientiousness after the experience In their theoretical review, Neyer et al. (2014) stated that of this event. Research did not provide evidence for the normativeness of a life event (or, in this case, a relation- socialization effects of widowhood on changes in ship event) is crucial when studying selection and social- neuroticism, extraversion, agreeableness, and openness. ization effects, for two reasons. First, the less normative an event is, the more likely personality contributes to the event (relevant for selection effects). Second, the more normative 1.2.3 | Interim summary on selection an event is, the more transparent are the associated role de- effects and socialization effects mands and behavioral scripts and the clearer is the guid- ance of how to behave adaptively (relevant for socialization Most of the studies reviewed used large nationally rep- effects; see also Caspi & Moffitt, 1993). This normativeness, resentative samples with long time lags between assess- in turn, facilitates adapting to a new status, which then ments (i.e., 1 to 22 years). These previous findings suggest guides personality change (Neyer et al.,  2014). Although that relationship events are transactionally linked with events are often too complex to be categorized in either of personality traits, but the size of the effects tended to be two categories (i.e., normative vs. non- normative), they can small (see Table  S1). Moreover, some inconsistency ex- be described along a continuum of normativeness accord- ists as to how relationship events are linked to personal- ing to three aspects (Neyer et al., 2014). ity traits. We argue these inconsistencies might have been First, normativeness depends on whether an event is con- driven by unnoticed moderators, namely by participants' sidered mandatory for people in a (sub)cultural context in a age. Many of the studies reviewed were either based on certain life period (Neyer et al., 2014). For example, marriage 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BÜHLER et al. | 5 is considered more mandatory in some cultures than in other are less normative the younger people are). Selection ef- cultures (e.g., the crude marriage rate, defined as the number fects on gain- based relationship events, conversely, should of marriages during the year per 1000 persons, is higher in be stronger the older people are (because gains are less Germany [5.0] than in Italy [3.1]; OECD, 2019). normative the older people are). Second, normativeness depends on whether an event Socialization effects of relationship events on person- is common and frequently experienced among members of ality traits should be stronger when the event is more nor- a reference group (i.e., peers of the same age in a [sub] mative because more normative events provide clearer population; Neyer et al., 2014). For example, widowhood advice and guidance about how to behave adaptively is more frequently experienced among adults between (Roberts et al., 2005). Accordingly, socialization effects of 65 and 74 years (from a total of 92,522 people, 20.7% in- gain- based events should be stronger the younger people dicated that they were currently widowed) than among are (because gains are more normative the younger people adults between 35 and 44 years (1.1% indicated current are). Socialization effects of loss-b ased events, conversely, widowhood) (U.S. Census Bureau, 2021). should be stronger the older people are (because losses are Third, normativeness depends on whether an event more normative the older people are). is socially expected and socially scripted among members of a reference group (Neyer et al., 2014). The normative timing of social expectations and scripts is, among other 1.4 | The present study aspects, defined by culture- specific social clocks (Bleidorn et al.,  2013; Neugarten et al.,  1965) and developmental People experience relationship events across entire adult- tasks (Havighurst, 1972). For example, young adults of re- hood, and evidence suggests that relationship events cent cohorts, compared to previous cohorts, may follow are transactionally linked with personality traits. So far, more diverse romantic scripts (Bühler & Nikitin,  2020; however, research has led to partly conflicting findings Scheling & Richter, 2021). as to how personality traits and relationship events pre- Following this developmental approach, we argue dict each other. In the case of such inconsistency, two ap- that age is crucial to determine the normativeness of proaches are needed. relationship events and expect that the strength of First, to examine systematic variations in selection transactions varies systematically across adulthood. and socialization effects, moderator variables need to be Specifically, according to developmental theories, young identified. We contend that progress in understanding adults (18– 40 years) typically focus on gains and growth personality– relationship transactions benefits from con- (Havighurst, 1972), establish long- lasting social ties, and sidering the role of age for transactions, based on the de- commit to their first long- term romantic relationships velopmental gains and losses that accompany the event. (Ebner et al., 2006). Middle- aged (40– 65 years) and older Second, to address the heterogeneity of previous (age 65 and above) adults, conversely, are more concerned findings, hypotheses should be tested in more than one with consolidation and avoiding losses, which is expressed data set, and evidence should be systematically sum- in caring for the next generation and in maintaining so- marized (Curran & Hussong, 2009; Duncan et al., 2014; cial relationships, including marriage (e.g., Baltes,  1987; Hofer & Piccinin,  2009). To that aim, we use longitu- Infurna et al.,  2020). Continuing with this developmen- dinal data from three nationally representative sam- tal view, this implies that gain-b ased relationship events ples: the German Socio-E conomic Panel (SOEP), the should be more common (and hence more normative) Household Income and Labour Dynamics in Australia for young adults, while loss- based relationship events (HILDA) Survey, and Understanding Society from the should be more common (and hence more normative) United Kingdom (Understanding Society is the UK for middle- aged and older adults. Thus, if normativeness Household Longitudinal Study, using waves of the plays a crucial role for transactions (Neyer et al., 2014) and British Household Panel Study [BHPS], harmonized if normativeness depends on age (e.g., Havighurst, 1972; with Understanding Society. BHPS is the household- Rubin et al., 2009) then selection and socialization effects based panel survey of residents of the UK, which ran should differ across adulthood in the following ways. from 1991 to 2009). We will first conduct study- level Selection effects of personality traits should be stronger analyses in each data set. Next, we will aggregate these when the event is less normative because less normative findings meta-a nalytically to increase the power of tests, events are less regulated by social expectations, which the precision of estimates, and the generalizability of gives personality more chance to shape the occurrence findings (Viechtbauer,  2005, 2010), allowing for inter- of the event (Neyer et al., 2014). Following this argumen- nal replications and robustness checks (e.g., Duncan tation, selection effects on loss- based relationship events et al., 2014). By pursuing these approaches, the present should be stronger the younger people are (because losses coordinated analysis offers the unique opportunity to 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 6 | BÜHLER et al. provide a clearer picture of the transactions between the research ethics officer from the German Institute for personality traits and relationships events and, most im- Economic Research. SOEP is an ongoing yearly household- portantly, to gain a better understanding of the role of based panel study of people above age 17 years living in age for these transactions. Germany (for more details, see Goebel et al., 2019; Wagner et al., 2007).3 Since 1985, relationship events have been as- sessed yearly, and since 2005, personality traits have been 2 | METHOD assessed every four years with the most recent assessment in 2017. Given our focus on traits, we used SOEP data from 2.1 | Transparency and openness all personality assessments (i.e., 2005 to 2017). We included individuals who had completed at least two of three items We follow the Journal Article Reporting Standards to assess traits on at least two consecutive assessments. The (Appelbaum et al., 2018; Kazak, 2018) and describe how event- specific samples are reported in Table 1 (left part). we obtained the three samples included in this study. The data sets are publicly available, and links for obtaining HILDA information on study protocol, data access, and publica- The second data set comes from HILDA of the Melbourne tions using these data sets are provided in Footnotes 1, 2, Institute of Applied Economic and Social Research, and 3. Analysis scripts and research materials (e.g., coding and ethical approval was given by the University of manual) are available on the Open Science Framework Melbourne's research ethics committee. HILDA is an on- (OSF; https://osf.io/4cxhz/). The present work was ex- going household-b ased panel survey of people above age plorative, and hypotheses and analyses were not preregis- 15 years living in Australia (for more details, see Melbourne tered. The analyses were computed in R (R Development Institute of Applied Economic and Social Research, 2017).4 Core Team, 2020), using the lavaan (Rosseel, 2012) and Since 2002, relationship events have been assessed yearly, metafor (Viechtbauer, 2010) packages. and since 2005, personality traits have been assessed every four years with the most recent assessment in 2017. We used HILDA data from 2005 to 2017 and included individu- 2.2 | Samples and procedures als who had completed at least two of the personality items on at least two consecutive assessments. The event-s pecific 2.2.1 | Three household- panel studies samples are reported in Table 1 (middle part). SOEP Understanding Society The first data set comes from SOEP of the German The third data set comes from Understanding Society of Institute of Economic Research and was approved by the University of Essex and was approved by the University T A B L E 1 Occurrence of relationship events and age at occurrence in the three data sets (SOEP, HILDA, Understanding society) SOEP HILDA Understanding Society Occurrence Age Occurrence Age Occurrence Age Variable Yes No M SD Yes No M SD Yes No M SD Gain-b ased event New relationship 3586 17,600 32.13 13.02 492 2812 34.13 13.30 – – – – Moving in 3959 16,453 32.69 11.55 3531 11,324 29.44 10.37 933 10,338 33.08 12.46 Marriage 3261 16,487 36.05 10.87 3552 10,569 35.10 12.83 1132 3075 36.78 11.91 Loss- based event Separation 3099 16,697 36.77 11.93 3982 10,450 36.57 14.94 348 3104 42.13 11.09 Divorce 976 17,880 43.39 9.46 808 12,272 43.91 11.01 555 2997 46.83 12.83 Widowhood 951 17,629 68.16 12.41 528 12,562 70.85 13.11 441 3066 69.66 12.44 Note: Occurrence indicates the number of people who have experienced the event at least once during the study period (only events are considered for which it can be ensured that the event took place between two consecutive personality measurements). Yes = Event occurred; No = Event did not occur. Age = Age at the occurrence of the event. In the SOEP data set, events were considered between 2005 to 2017, except for beginning a relationship, which was assessed between 2011 and 2017. In the HILDA data set, events were considered between 2005 to 2017. In the data set of Understanding Society, events were considered between 2005 and 2011, except for moving in with a partner, which was assessed between 2010 and 2011. The event “beginning a relationship” was not assessed in the data set of Understanding Society. 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BÜHLER et al. | 7 of Essex ethics committee.5 Since 2009, Understanding and HILDA data sets. For each relationship event, we di- Society has been conducting yearly interviews with chotomously coded whether participants experienced the around 40,000 households, including 8000 of the original event at least once during the study period (coded with BHPS households. Since 1992, relationship events have “1”) or did not experience the event during the study pe- been assessed yearly. Personality traits have been assessed riod (coded with “0”). Thus, the resulting relationship- twice (2005 and 2011). In this research, we used data from event variable contrasted between those participants individuals who had completed at least two of the person- who experienced the event (i.e., event sample) and those ality items at both assessments. The event- specific sam- participants who did not experience the event during the ples are reported in Table 1 (right part). specific time period (i.e., control sample). If participants had experienced a relationship event more than once, we used data from the first occurrence. We set this focus be- 2.3 | Measures cause repeated events tend to be more frequent for gain- based events than for loss-b ased events, suggesting that 2.3.1 | Personality traits considering multiple events might bias the conclusions (Denissen et al., 2019; Luhmann & Eid, 2009). Below, we describe the measures of personality traits in each data set. The event-s pecific Cronbach's alphas are reported in Table  S2 and ranged between 0.48 to 0.82. 2.4 | Data- analysis approach The relatively low internal consistencies have also been reported in previous research using these nationally rep- To test the associations between personality traits and re- resentative data sets (Dyrenforth et al., 2010). lationship events, we used latent growth models (LGMs; Bollen & Curran, 2006; Grimm et al., 2016). LGMs are well SOEP suited to study overall change in personality traits (e.g., Personality traits were assessed with the Big Five Jackson & Allemand, 2014) and allow testing selection ef- Inventory-S OEP (BFI- S; Schupp & Gerlitz,  2014), based fects and socialization effects (see Specht et al., 2011). To on the Big Five Inventory (John et al.,  1991). Each Big test these effects, we had to restructure the data in the fol- Five trait was assessed with three items on a 7- point scale lowing ways (see Figures S2 and S3). (1 = “not at all” to 7 = “absolutely”). In the event sample, we restructured the personality data depending on the relationship event. More precisely, HILDA we used data from the last personality assessment before Personality traits were assessed with an adaptation of the event and from the first personality assessment after the 36- item version of the Trait Descriptive Adjectives the event, which ensured temporal proximity between (Saucier, 1994). Neuroticism, extraversion, conscientious- traits and events. In the remainder, we refer to the pre- ness, and openness were measured with six items, while event and post- event personality assessments as Times 1 agreeableness was measured with four items. Responses and 2, respectively. In the control sample, we had to ensure were assessed on a 7-p oint scale (1 = “does not describe that the time lag between personality assessments was me at all” to 7 = “describes me very well”). identical to the time lag in the event sample. Therefore, we used data from two consecutive personality assessments Understanding Society that were randomly chosen and equally balanced over the Personality traits were measured with a 15-i tem version study period. We refer to these personality assessments as of the Big Five Inventory (John & Srivastava, 1999). Each Times 1 and 2. We used first-o rder models, in which in- trait was assessed with three items on a 7- point scale tercepts and slopes were modeled with manifest indicator (1 = “does not apply to me at all” to 7 = “applies to me variables from Times 1 and 2 (see Figure S1). perfectly”). We note that using household data means that some participants were clustered in households. Tables  S3– S5 show the percentages of participants with the same 2.3.2 | Relationship events household ID in the event (left part) and control (right part) samples. As the tables indicate, percentages were be- In each data set, we coded three gain-b ased relationship tween 0% and 24% in the event samples and between 6% events (i.e., beginning a relationship, moving in with part- and 22% in the control samples, suggesting that around ner, marriage) and three loss- based relationship events 80% were independent data. Hence, we did not expect sys- (i.e., separation, divorce, widowhood).6 The event of be- tematic effects of household clustering on selection and ginning a relationship was included only in the SOEP socialization effects. Moreover, although there exists the 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 8 | BÜHLER et al. lavaan.survey package (Oberski, 2014) to deal with clus- of personality change across assessments. Second, we tered data, the interaction terms with age, which were included interaction terms between the relationship crucial in the present study, cannot be modeled with this event and the linear and quadratic age variables and package. Hence, we conducted the analyses without con- regressed these terms on both the intercept and the trolling for household ID. slope. A significant interaction effect on the intercept Overall, our approach (a) allowed using all available indicated that people with higher (or younger) age have information on relationship events, (b) ensured temporal a stronger (or weaker) selection effect, while a signifi- proximity between traits and events in the event sample, cant interaction effect on the slope indicated that peo- and (c) considered data with identical time lags in the ple with higher (or younger) age have a stronger (or event and control samples, which substantially increased weaker) socialization effect. For each event, we used the reliability of the event measure and the power for the an event- specific age variable: In the event sample, we analyses. Furthermore, the use of LGMs enabled us to used the average age at which participants experienced study both personality change that was independent of the specific event, and in the control sample we used relationship events and personality change that was pre- participants' mean age at Time 1. Age was grand-m ean dicted by relationship events. To deal with missing values, centered and, to avoid numerically small estimates, we used listwise deletion, which is the default behavior in rescaled by the factor 10. the lavaan package if data include missing values. The sig- nificance level was set to p < 0.01 due to the considerable number of tests. Gender was entered as covariate on the 2.5 | Measurement invariance intercept and slope in all models. Given that the scores of personality traits are comparable over time only if factorial invariance is given (Widaman 2.4.1 | Operationalization of selection effects, et al., 2010), we tested for measurement invariance across socialization effects, and age effects assessments (i.e., pre-e vent and post-e vent assessments). We tested three measurement models per trait and event: A selection effect was operationalized as the effect of the Model 1 included configural invariance for the indicator relationship event on the intercept of the personality trait variables. Model 2 tested metric invariance by constrain- (see also Specht et al., 2011). The path coefficients of the ing the loadings to be equal across assessments and Model intercept were constrained to 1 across both assessments. A 3 tested scalar invariance by constraining thresholds to significant effect of the relationship event on the intercept be equal across assessments. To allow for pre-e vent and indicated that participants who experienced (vs. did not post- event comparisons of mean levels, we had to ensure experience) the event had a lower (or higher) score in the that measures showed scalar invariance. To assess model trait measure at Time 1. In other words, the Time 1 trait fits, we used the comparative fit index (CFI) and the root predicted the occurrence of the event between Time 1 and mean square error of approximation (RMSEA) with CFI Time 2. ≥ 0.95 and RMSEA ≤ 0.06 indicating a good model fit (Hu A socialization effect was operationalized as the effect & Bentler, 1999). We considered a change of ≤0.01 in CFI of the relationship event on the slope of the personality as indicative of measurement invariance (Chen,  2007; trait (see also Specht et al., 2011). The path coefficients of Cheung & Rensvold, 2002). the slope were constrained to 0 (i.e., Time 1) and 1 (i.e., Time 2). A significant effect of the event on the slope indi- cated that participants who experienced (vs. did not expe- 3 | RESULTS rience) the event differed in their average rate of change in the personality trait across assessments. Thus, the occur- 3.1 | Descriptive information and rence of a relationship event predicted personality change preliminary analyses between Time 1 and Time 2. We included age effects into the models in two ways. Table  1 provides an overview of the occurrence of rela- First, we regressed both the intercept and the slope tionship events in each data set, including participants' on a linear and quadratic age variable. A significant mean age at the occurrence of the event. Relationship age effect on the intercept indicated that people with events were experienced by around 5% to 25% of partici- higher (or younger) age had a higher (or lower) Time pants in each sample during the respective study period, 1 mean value of the personality trait. A significant age with lowest occurrence rates for widowhood and highest effect on the slope indicated that people with higher occurrence rates for marriage and separation. On average, (or younger) age experienced a steeper (or flatter) rate participants tended to be younger when experiencing a 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BÜHLER et al. | 9 gain- based relationship event, compared to a loss-b ased 3.2 | Latent growth models relationship event, and participants' mean ages were closer together in the case of gain- based events (29.44 The LGMs were fully saturated. Below, we first report se- to 36.78 years), compared to loss-b ased events (36.57 to lection and socialization effects, which are summarized in 70.85 years). Figures 1 and 2, and then discuss the effects of age (for all Tables S3–S 6 report descriptive information on person- details, see Tables S27–S 43). ality traits, ordered by gain- based and loss- based events in the event and control samples of each data set. As the tables indicate, mean- level changes in personality traits 3.2.1 | Selection and socialization effects emerged in both the event and control samples. In the event samples, fewer mean- level changes were observed Figure 1 depicts the point estimates and 99% confidence when people experienced the loss-b ased relationship intervals for selection effects in each data set. Regarding events of separation and marriage. Tables S7–S 23 report gain-b ased events, (a) higher extraversion and higher the fit indices of the measurement models to test invari- openness were significantly linked with beginning a re- ance. As the tables indicate, the fits of Model 3 (i.e., scalar lationship, (b) higher extraversion and higher openness measurement invariance) were generally good, indicating were significantly linked with moving in with the partner, that mean levels could be compared across the pre- event and (c) higher extraversion, higher openness, and higher and post- event assessments. agreeableness were significantly linked with marrying. Summary of Standardized Coefficients for Selection Effects of Latent Growth Curve Models Neuroticism Extraversion Agreeableness Conscientiousness Openness SOEP −0.02 [−0.06; 0.03] 0.05 [0.01; 0.09] 0.01 [−0.03; 0.05] 0.01 [−0.04; 0.05] 0.05 [0.01; 0.09] HILDA −0.02 [−0.09; 0.06] 0.08 [−0.01; 0.16] 0.03 [−0.05; 0.11] 0.01 [−0.08; 0.10] 0.06 [−0.03; 0.15] Understanding Society SOEP −0.01 [−0.05; 0.03] 0.04 [0.00; 0.08] 0.01 [−0.03; 0.05] 0.03 [−0.01; 0.07] 0.03 [−0.00; 0.07] HILDA −0.04 [−0.09; 0.01] 0.03 [−0.02; 0.07] 0.01 [−0.04; 0.05] −0.00 [−0.05; 0.05] 0.04 [−0.00; 0.08] Understanding Society −0.06 [−0.14; 0.03] 0.08 [−0.02; 0.18] 0.03 [−0.07; 0.13] 0.01 [−0.10; 0.13] 0.12 [0.03; 0.21] SOEP −0.02 [−0.06; 0.02] 0.05 [0.01; 0.09] −0.03 [−0.07; 0.01] −0.01 [−0.05; 0.03] 0.03 [−0.01; 0.07] HILDA −0.03 [−0.08; 0.02] 0.01 [−0.03; 0.06] 0.05 [0.00; 0.09] 0.01 [−0.03; 0.06] 0.05 [0.00; 0.09] Understanding Society −0.04 [−0.11; 0.04] 0.03 [−0.05; 0.10] 0.04 [−0.04; 0.11] 0.01 [−0.07; 0.09] 0.06 [−0.01; 0.13] SOEP 0.01 [−0.03; 0.04] 0.04 [0.00; 0.07] −0.03 [−0.06; 0.01] 0.00 [−0.03; 0.04] 0.04 [0.01; 0.07] HILDA −0.09 [−0.13; −0.05] 0.03 [−0.01; 0.07] −0.02 [−0.06; 0.02] −0.04 [−0.08; −0.00] 0.05 [0.01; 0.09] Understanding Society −0.01 [−0.08; 0.06] 0.04 [−0.03; 0.12] 0.02 [−0.05; 0.09] 0.03 [−0.05; 0.12] −0.01 [−0.09; 0.07] SOEP 0.01 [−0.02; 0.04] 0.02 [−0.01; 0.06] 0.00 [−0.03; 0.03] −0.01 [−0.04; 0.02] 0.02 [−0.02; 0.05] HILDA −0.00 [−0.04; 0.03] 0.03 [−0.01; 0.07] 0.02 [−0.01; 0.06] 0.01 [−0.03; 0.05] 0.04 [0.00; 0.08] Understanding Society −0.00 [−0.07; 0.07] 0.04 [−0.03; 0.11] 0.02 [−0.04; 0.09] 0.00 [−0.07; 0.07] 0.03 [−0.04; 0.09] SOEP 0.01 [−0.04; 0.06] −0.02 [−0.07; 0.03] −0.01 [−0.07; 0.05] 0.02 [−0.03; 0.07] −0.04 [−0.10; 0.01] HILDA −0.06 [−0.15; 0.03] −0.01 [−0.08; 0.07] 0.03 [−0.05; 0.10] −0.02 [−0.10; 0.06] 0.01 [−0.06; 0.08] Understanding Society −0.03 [−0.12; 0.06] 0.06 [−0.05; 0.17] 0.01 [−0.11; 0.13] 0.04 [−0.06; 0.14] 0.01 [−0.09; 0.11] −0.25 0 0.25 −0.25 0 0.25 −0.25 0 0.25 −0.25 0 0.25 −0.25 0 0.25 Standardized Parameter Estimate F I G U R E 1 Summary of standardized coefficients for selection effects of latent growth curve models. Note. The figure shows standardized estimates and their 99% confidence intervals. The event “beginning a relationship” was not assessed in the data set of Understanding Society. 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License New Relationship Moving in Marriage Separation Divorce Widowhood 10 | BÜHLER et al. Summary of Standardized Coefficients for Socialization Effects of Latent Growth Curve Models Neuroticism Extraversion Agreeableness Conscientiousness Openness SOEP −0.02 [−0.07; 0.02] −0.01 [−0.05; 0.03] −0.01 [−0.05; 0.03] 0.01 [−0.03; 0.05] −0.01 [−0.05; 0.03] HILDA −0.00 [−0.09; 0.09] −0.05 [−0.13; 0.04] −0.04 [−0.11; 0.04] −0.00 [−0.09; 0.08] 0.01 [−0.06; 0.09] Understanding Society SOEP 0.02 [−0.03; 0.06] −0.01 [−0.06; 0.03] −0.02 [−0.06; 0.02] −0.03 [−0.07; 0.01] −0.01 [−0.05; 0.03] HILDA 0.01 [−0.05; 0.06] −0.00 [−0.05; 0.04] 0.03 [−0.01; 0.06] 0.01 [−0.04; 0.06] −0.01 [−0.05; 0.03] Understanding Society 0.04 [−0.06; 0.14] 0.01 [−0.06; 0.09] −0.03 [−0.14; 0.07] −0.02 [−0.11; 0.08] −0.02 [−0.11; 0.08] SOEP 0.01 [−0.03; 0.05] −0.04 [−0.08; −0.00] −0.00 [−0.04; 0.04] 0.00 [−0.04; 0.04] −0.02 [−0.06; 0.02] HILDA −0.01 [−0.06; 0.05] −0.01 [−0.06; 0.04] −0.02 [−0.06; 0.03] −0.01 [−0.06; 0.04] −0.02 [−0.06; 0.03] Understanding Society 0.03 [−0.04; 0.11] 0.01 [−0.07; 0.09] −0.05 [−0.13; 0.03] −0.00 [−0.08; 0.07] −0.04 [−0.11; 0.03] SOEP −0.02 [−0.06; 0.01] 0.01 [−0.03; 0.04] 0.05 [0.01; 0.08] −0.01 [−0.04; 0.03] 0.02 [−0.02; 0.05] HILDA 0.03 [−0.02; 0.07] −0.01 [−0.05; 0.03] 0.03 [−0.01; 0.07] −0.03 [−0.06; 0.01] 0.03 [−0.01; 0.07] Understanding Society 0.00 [−0.08; 0.09] 0.03 [−0.05; 0.10] −0.02 [−0.09; 0.06] 0.00 [−0.06; 0.07] 0.03 [−0.05; 0.10] SOEP −0.02 [−0.05; 0.02] −0.01 [−0.05; 0.02] 0.01 [−0.02; 0.04] 0.00 [−0.03; 0.03] 0.01 [−0.02; 0.04] HILDA 0.01 [−0.04; 0.05] 0.02 [−0.02; 0.06] −0.01 [−0.04; 0.02] 0.01 [−0.03; 0.04] −0.00 [−0.04; 0.03] Understanding Society −0.02 [−0.09; 0.06] −0.02 [−0.09; 0.05] −0.02 [−0.09; 0.06] −0.02 [−0.09; 0.06] −0.03 [−0.10; 0.04] SOEP 0.01 [−0.05; 0.07] −0.01 [−0.06; 0.04] 0.01 [−0.05; 0.06] −0.01 [−0.06; 0.05] −0.00 [−0.06; 0.06] HILDA 0.02 [−0.07; 0.11] −0.02 [−0.11; 0.07] 0.03 [−0.06; 0.12] −0.04 [−0.11; 0.04] 0.03 [−0.06; 0.11] Understanding Society −0.07 [−0.18; 0.05] 0.03 [−0.07; 0.12] 0.05 [−0.04; 0.14] −0.00 [−0.09; 0.08] 0.00 [−0.11; 0.12] −0.25 0 0.25 −0.25 0 0.25 −0.25 0 0.25 −0.25 0 0.25 −0.25 0 0.25 Standardized Parameter Estimate F I G U R E 2 Summary of standardized coefficients for socialization effects of latent growth curve models. Note. The figure shows standardized estimates and their 99% confidence intervals. The event “beginning a relationship” was not assessed in the data set of Understanding Society. Regarding loss-b ased events, lower neuroticism, higher Next, we meta-a nalytically aggregated the findings extraversion, lower conscientiousness, and higher open- across data sets, to gain a more comprehensive picture ness were significantly linked with separation. No sig- on selection and socialization effects (Table 2). We note, nificant selection effects were observed on divorce and however, that meta- analytic computations based on few widowhood. Moreover, only one of the reported selection (even large) data sets may be underpowered, and conclu- effects (i.e., openness on separation) replicated across two sion should be drawn with caution. Therefore, the meta- datasets, and none of the effects replicated across all three analytic computations mainly serve illustrative purposes datasets. However, as the point estimates in Figure 1 illus- rather than providing robust effect size estimates (for a trate, the effects were often in similar ranges. similar approach, see Mund et al., 2020). Nevertheless, the Figure  2 shows the point estimates and 99% confi- data sets are among the largest and most representative dence intervals for socialization effects in each data set. data sets that can be used to study transactions, and thus The gain- based event marriage predicted significant de- the aggregated point estimates shown in Table 2 may serve creases in extraversion, while the loss-b ased event sep- as suitable estimates for future research in this field. The aration predicted significant increases in agreeableness. heterogeneity indices in the Table (Q, τ2, and I2) inform No significant socialization effects were observed for be- about the consistency of the estimates across data sets ginning a relationship, moving in with a partner, divorce, (Borenstein et al., 2017), but, again, need to be interpreted and widowhood, and none of the reported effects repli- with caution when the number of studies is small (Huedo- cated across data sets. However, the effects were again in Medina et al., 2006). In general, selection effects tended to similar ranges. be more consistent than socialization effects, indicated by 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License New Relationship Moving in Marriage Separation Divorce Widowhood BÜHLER et al. | 11 T A B L E 2 Meta- analytic aggregation of coefficients for selection and socialization effects across samples Neuroticism Extraversion Agreeableness Heterogeneity Heterogeneity Heterogeneity Model ES 99% CI Q τ2 I2 ES 99% CI Q τ2 I2 ES 99% CI Q τ2 I2 Selection effects New relationship −0.02 [−0.05, 0.01] 0.0 0.00 0.0 0.06 [0.02, 0.09] 1.4 0.00 29.6 0.01 [−0.01, 0.04] 0.6 0.00 0.0 Moving in −0.03 [−0.07, 0.02] 3.4 0.00 41.5 0.04 [0.02, 0.07] 2.1 0.00 4.8 0.01 [−0.01, 0.03] 0.5 0.00 0.0 Marriage −0.03 [−0.04, −0.01] 0.9 0.00 0.0 0.03 [−0.01, 0.07] 6.9 0.00 71.0 0.02 [−0.06, 0.09] 30.1 0.00 93.4 Separation −0.03 [−0.12, 0.06] 43.6 0.00 95.4 0.04 [0.02, 0.05] 0.5 0.00 0.0 −0.02 [−0.05, 0.01] 4.0 0.00 49.7 Divorce 0.004 [−0.01, 0.02] 0.8 0.00 0.0 0.03 [0.01, 0.04] 0.9 0.00 0.0 0.01 [−0.01, 0.03] 1.6 0.00 0.0 Widowhood −0.03 [−0.09, 0.04] 21.0 0.00 90.5 0.004 [−0.04, 0.05] 10.1 0.00 80.2 0.01 [−0.03, 0.05] 6.8 0.00 70.6 Socialization effects New relationship −0.02 [−0.04, 0.01] 0.5 0.00 0.0 −0.03 [−0.07, 0.02] 2.5 0.00 60.2 −0.02 [−0.05, 0.02] 1.4 0.00 29.3 Moving in 0.02 [−0.001, 0.04] 0.6 0.00 0.0 −0.01 [−0.03, 0.02] 0.5 0.00 0.0 −0.02 [−0.04, 0.004] 1.6 0.00 0.0 Marriage 0.01 [−0.02, 0.03] 3.1 0.00 36.1 −0.02 [−0.05, 0.02] 6.5 0.00 69.2 −0.02 [−0.05, 0.01] 4.7 0.00 57.3 Separation 0.004 [−0.04, 0.05] 10.7 0.00 81.3 0.01 [−0.02, 0.03] 3.0 0.00 33.6 0.03 [−0.02, 0.07] 8.2 0.00 75.5 Divorce −0.01 [−0.04, 0.02] 4.1 0.00 50.8 −0.001 [−0.03, 0.03] 4.6 0.00 56.8 −0.001 [−0.02, 0.02] 2.5 0.00 20.6 Widowhood −0.01 [−0.06, 0.04] 12.2 0.00 83.7 −0.01 [−0.04, 0.02] 3.7 0.00 45.3 0.02 [−0.002, 0.05] 3.4 0.00 41.4 Conscientiousness Openness Heterogeneity Heterogeneity Model ES 99% CI Q τ2 I2 ES 99% CI Q τ2 I2 Selection effects New relationship 0.01 [−0.02, 0.04] 0.0 0.00 0.0 0.06 [0.01, 0.10] 0.10 0.00 0.0 Moving in 0.03 [0.004, 0.05] 1.0 0.00 0.0 0.06 [−0.02, 0.15] 10.1 0.00 80.2 Marriage −0.001 [−0.02, 0.02] 2.0 0.00 0.0 0.04 [0.02, 0.06] 2.6 0.00 23.1 Separation −0.01 [−0.05, 0.04] 11.1 0.00 82.1 0.03 [0.001, 0.07] 5.3 0.00 62.2 Divorce −0.002 [−0.02, 0.02] 1.7 0.00 0.0 0.03 [0.01, 0.05] 1.7 0.00 0.0 Widowhood 0.01 [−0.03, 0.05] 8.9 0.00 77.6 −0.01 [−0.06, 0.04] 12.1 0.00 83.4 Socialization effects New relationship 0.01 [−0.02, 0.03] 0.2 0.00 0.0 0.00 [−0.04, 0.04] 0.4 0.00 0.0 Moving in −0.03 [−0.05, −0.004] 1.0 0.00 0.0 −0.01 [−0.04, 0.01] 0.1 0.00 0.0 Marriage −0.002 [−0.02, 0.02] 0.7 0.00 0.0 −0.02 [−0.04, −0.004] 0.7 0.00 0.0 (Continues) 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 12 | BÜHLER et al. somewhat higher estimates of I2 for socialization effects than for selection effects. 3.2.2 | Effects of age Tables S27– S43 report the effects of age, indicating sig- nificant main effects of age on personality traits. Most consistently, participants reported higher levels in agreeableness and conscientiousness with higher age. Yet, few significant age moderations emerged. More precisely, we observed significant interactions between age and relationship events with selection and sociali- zation effects in six cases. Regarding selection effects, we observed significant interactions between (a) linear age and moving in with a partner in the analyses with conscientiousness in the SOEP data set (β =  0.06, 99% CI [0.03, 0.10]), (b) linear age and marriage in the analy- ses with neuroticism in the HILDA data set (β = −0.05, 99% CI [−0.09, −0.01]), (c) linear age and marriage in the analyses with extraversion in the HILDA data set (β  =  −0.04, 99% CI [−0.08, −0.004]), and linear age and widowhood in the analyses with extraversion in the HILDA data set (β  =  −0.19, 99% CI [0.01, 0.37]). Regarding socialization effects, we observed significant interactions between (d) linear age and beginning a re- lationship in the analyses with conscientiousness in the HILDA data set (β = −0.08, 99% CI [−0.15, −0.01]), and (e) quadratic age and moving in with a partner in the analyses with neuroticism in the Understanding Society data set (β = −0.08, 99% CI [−0.15, −0.002]). Hence, the great majority of age moderations occurred with gain- based relationship events. 4 | DISCUSSION In this study, we examined selection and socialization effects between personality traits and relationship events in three nationally representative data sets. We classified relation- ship events into gain- based and loss- based events, which al- lowed us to test theoretically derived predictions about how age would matter for personality–r elationship transactions. Overall, the findings indicated stronger selection than so- cialization effects (mainly with gain- based events) with only few country- specific age- moderation effects. 4.1 | Selection effects rather than socialization effects So far, evidence was partly mixed as to how personality traits and relationship events are transactionally linked T A B L E 2 (Continued) Conscientiousness Openness Heterogeneity Heterogeneity Model ES 99% CI Q τ2 I2 ES 99% CI Q τ2 I2 Separation −0.02 [−0.04, 0.01] 2.4 0.00 18.1 0.02 [0.01, 0.04] 0.5 0.00 0.0 Divorce 0.002 [−0.02, 0.02] 1.4 0.00 0.0 0.001 [−0.02, 0.02] 2.6 0.00 24.0 Widowhood −0.02 [−0.05, 0.01] 4.4 0.00 54.1 0.01 [−0.02, 0.04] 4.2 0.00 51.8 Note: Computations were made with random-e ffects models. The number of samples was 3 for all models except for beginning a relationship (k = 2). ES = Weighted mean effect size, indicating the effect of personality traits on the occurrence of relationship events (for selection effects) or the effect of relationship events on change in personality traits (for socialization effects). CI = confidence interval; Q = statistic used in heterogeneity test; τ2 = estimated amount of total heterogeneity; I2 = ratio of total heterogeneity to total variability (given in percent). Values in bold are significant at p < 0.01. 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BÜHLER et al. | 13 with each other (Bleidorn et al.,  2018). In the present The finding that no consistent socialization effects study, we sought to provide a more robust picture by test- emerged corresponds with previous findings, showing ing transactions between personality traits and relation- that people do not— as would be expected theoretically ship events in three household panel data sets. Across five (Roberts & Wood,  2006)—b ecome psychologically more personality traits and six relationship events, the overall mature, when they transition into a new social role, picture is that transactions occurred rarely and that the such as becoming a mother or father (van Scheppingen size of the observed effects was rather small. Selection et al., 2016). Although their study focused on parenthood, effects tended to be stronger and more consistent than which was not among the relationship events included in socialization effects, which corresponds with previous this study due to the age- specificity of this event, both the research (e.g., Asendorpf & Wilpers,  1998; Denissen study by van Scheppingen et al.  (2016) and the present et al., 2019; Neyer & Asendorpf, 2001; Specht et al., 2011) work revealed that transitioning into a new social role does and may suggest that relationships are more sensitive to not necessarily trigger psychological maturity. There are, personality effects than vice versa. at least, two reasons for why socialization effects emerge The selection effects that we observed occurred more rarely and inconsistently: First, relationship events do frequently with gain-b ased (vs. loss-b ased) events, and the effectively not change personality. Second, relationship two traits that were most dominant in this regard were events may change personality, but a more fine- grained extraversion and openness. In general, people higher in analysis of the mechanisms and conditions is needed to extraversion or openness were more likely to begin a re- understand when and how personality changes in re- lationship, to move in with a partner, to marry, and to sponse to relationship events. For instance, the inconsis- separate. Openness was also the only trait that showed con- tency observed for socialization effects may indicate that sistent effects across data sets (i.e., selection effects on sep- people do not consistently react to relationship events but aration; see Figure 1 and Tables S32 and S33). In general, vary in their reactions. In other words, people may show people higher in extraversion are more sensitive to reward- greater individual differences in response to a relationship ing stimuli and tend to have higher levels of energy, dom- event, compared to when they select themselves into a re- inance, and positivity (Smilie, 2013; Wilt & Revelle, 2017). lationship event, and more knowledge is needed to better People higher in openness are more motivated to approach understand these individual variations. and to create new experiences, tend to have higher levels of intelligence, curiosity, and creativity, and are characterized by being more open to change (DeYoung, 2015; McCrae & 4.2 | How does age matter? Costa Jr., 1997). While extraversion has been consistently linked with social interactions and relationships (Harris We observed significant main effects of age on personality & Vazire, 2016; Wrzus & Neyer, 2016), openness has often traits, suggesting that people develop as they age, corre- been considered an intellectual rather than “social” trait sponding with previous research (e.g., Caspi et al., 2005; (John & Srivastava,  1999). The current findings demon- Roberts et al., 2006). The present data, however, indicated strate that both traits, extraversion and openness, are rel- few significant interactions between age and relationship evant for individual differences in social relationships, as events with selection and socialization effects. Hence, in- they enable people to initiate and to engage in new expe- dividuals develop differently strongly across the life span, riences, including new experiences in the romantic rela- but individual differences in personality development can- tionship domain. The predominance of extraversion and not validly be captured by the occurrence of relationship openness regarding selection effects corresponds with events. This also indicates that personality development Digman's  (1997) higher- order factor β. Subsuming extra- was not more or less pronounced depending on whether version and openness (or intellect), factor β can be inter- individuals experienced a certain relationship event at a preted as personal growth and self- actualization, involving given age or not. The six significant age interactions that exploration of social and intellectual domains. It is likely we observed occurred mainly for gain- based relationship that a strong factor β (i.e., indicated by high values in extra- events, which could speak for a greater importance of version and openness) motivates people to approach and to gain- based relationship events when interacting with age. select new experiences, including relationship experiences. Nevertheless, a central conclusion from the findings is Moreover, a strong factor β contains an openness towards that personality– relationship transactions occurred rela- all life experiences, including the risky and potentially neg- tively independent of age. ative experiences, such as a loss-b ased relationship event Consistent with previous literature, we classified re- (Digman, 1997). This, in turn, may explain why the most lationship events into gain-b ased and loss- based events robust selection effects were found for a loss-b ased rela- (Denissen et al.,  2019). Still, this twofold distinction is tionship event (i.e., openness on separation). rather broad given that each relationship event itself 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 14 | BÜHLER et al. includes aspects of gains and losses. For example, be- 4.3 | The relevance of broader and ginning a relationship may fall into the category of gain- narrower environmental influences based relationship events, but may be accompanied by a series of individual gains and losses (e.g., gaining intimacy In the following, we discuss four sets of influences that and closeness, but losing flexibility and independence). may be relevant for how people select and respond to rela- Similarly, the presumably loss-b ased relationship event of tionship events and that may contribute to explaining the separation might involve both losses and gains (e.g., loos- strength of transactions across adulthood. ing existing closeness, but gaining new independence). Thus, the categorization and stratification of relationship events might be more complex than a dichotomy allows to 4.3.1 | Social scripts capture (Luhmann et al., 2020). In addition, we conceptualized normativeness based A first set that may impact the transactions between per- on a developmental approach, that is, depending on sonality traits and relationship events are the societal whether gains and losses accompany the relationship scripts that are associated with the relationship event. event. However, other conceptualizations of normative- Scripts differ between cohorts and cultures (Bleidorn ness are tenable, such as the mere frequency of rela- et al., 2013) and they navigate people through their love tionship events in certain developmental periods (e.g., lives (Dunlop et al., 2017). Deviations from these scripts Mayer, 2009). For example, in young adulthood individ- and differences in individual commitments to these uals often try different relationships and may leave un- scripts may account for individual differences in selec- satisfying relationships more readily, which could make tion and socialization effects. For instance, someone separation fairly normative in young adulthood, sim- who strongly commits to the step of marrying may show ply because it is part of an explorative paradigm (e.g., stronger socialization effects than someone who steps into Arnett, 2000). marriage out of societal sense of duty or external pressure. Finally, we used people's chronological age to test as- In fact, cross-s ectional data from a meta-a nalysis on so- sociations with age. However, other age- related aspects, cial investment and personality traits demonstrated that such as the age difference between both partners or peo- personality change towards greater psychological matura- ple's subjective age, may also be important to understand tion (i.e., lower neuroticism, higher agreeableness, higher how personality–r elationship transactions unfold across conscientiousness; Roberts et al., 2006) is associated with adulthood. For instance, at the time of marriage, men cognitive and emotional investment in the social role are usually 2.5 years older than women (Statista, 2021), rather than with the pure change in status (Lodi-S mith which may consequently be considered a normative age & Roberts,  2007). Thus, rather than the mere exposure difference for marriage. If a given couple has a non- to relationship events, the psychological investment and normative age difference (e.g., the woman being 4 years commitment associated with the relationship event may older than the man), this could result in stronger (or be crucial for transactions. weaker) personality– relationship transactions. Also, partners may behave differently depending on the other partner's age. For example, a 55- year- old person, who is 4.3.2 | Idiosyncratic experience and meaning in a relationship with a 35- year- old partner, may rather focus on gains and gain-r elated situations, compared to a A second set that may matter for personality–r elationship 55- year- old person, who is in a relationship with a same- transactions are the idiosyncratic experiences of rela- aged or older partner. Finally, people also differ in how tionship events, that is, the individual meaning that peo- old they feel compared with their chronological age (i.e., ple ascribe to a relationship event, such as divorce (e.g., subjective age; Pinquart & Wahl, 2021), which, in turn, Bühler & Dunlop, 2019; Haehner et al., 2021; Luhmann may have implications for the strength of selection and et al., 2020). In other words, rather than the relationship socialization effects (Stephan et al., 2014). For example, event per se, it may be the perception and interpretation a 55-y ear-o ld person who feels like 40 years may rather of the event that may be associated with selection effects approach and select gain-b ased situations and may react and, particularly, with socialization effects. For example, differently to these situations than a 55- year- old person someone who sees a divorce as a possible, almost natural, who feels like 60 years. Therefore, as noted above, it is component of relationship trajectories may react differ- essential to broadening the scope and to discuss further ently to their own divorce than someone who sees a di- influences that may impact personality– relationship vorce as a personal, or relational, failure. These different transactions across adulthood. meanings, in turn, may bring the divorcee into a distinct 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BÜHLER et al. | 15 role and will likely guide them differently through the fol- relationship events and personality assessments, for ex- lowing months and years, which, in turn, has implications ample through assessing micro- events that are happening for their feelings, thoughts, and behaviors (Lodi- Smith over the course of the relationship (Bleidorn et al., 2018). & Roberts, 2007; Roberts & Wood, 2006). Therefore, be- Moreover, less is known about the specific daily expe- ginning a relationship, moving in together, and marry- riences that people have immediately after a relationship ing may individually not be perceived as a gain but as a event. For instance, couple members who move into a cultural, familial, or social obligation. On the contrary, shared household need to adapt to the new affordances separation and divorce may individually not be perceived of a shared home and must negotiate their responsibili- as a loss but as a liberation. Even the sorrowful event of ties. Therefore, people who experience a relationship widowhood may be perceived as a relief for the deceased event likely create new daily situations, expectations, and partner. Hence, putting more emphasis on the meaning of behaviors, which may lead to changes in their daily lives events in an individual's life course is an important path and to long-t erm changes in their personality (Wrzus & for future research (see also Denissen et al., 2019). Roberts, 2017). For instance, recent findings have demon- strated that people who show more affective reactivity to daily hassles, such as conflicts, increase in their neuroti- 4.3.3 | Quality of the relationship cism over six years (Wrzus et al., 2021). Hence, to better understand the potential impact of relationship events, A third set that may matter for personality– relationship such as moving in together, on personality traits, it is cru- transactions is the quality of the relationship experience. cial to zoom into couples' daily lives, with its potential For instance, knowing that a person began a new relation- daily stressors, to examine the mechanisms that lead to ship says little about how satisfied this person is with the long- term personality development. Moreover, the daily new relationship. Similarly, losing a relationship that was experiences that people encounter may differ depending low in quality may have a different impact than losing a on the personality of their partner. For example, moving relationship that was high in quality and closely related in with an organized and structured partner (i.e., high in to the person's self-c oncept (Lewandowski Jr. et al., 2006). conscientiousness) may create different daily situations From research on self- esteem, we know that the quality of and routines than living together with a less organized the relationship may explain both selection effects on sepa- and less structured partner (i.e., low in conscientious- ration and socialization effects of beginning a relationship ness), which, in turn, may lead to stronger (or weaker) (Luciano & Orth,  2017). Moreover, satisfaction with the socialization effects. Hence, future research is needed that relationship and the event also depends on characteristics studies both partners' daily experiences close to a relation- of the partner and how he or she reacts to the event (e.g., ship event (Wrzus & Roberts, 2017). Dyrenforth et al., 2010), further contributing to the com- To conclude, people may select themselves into cer- plexity of personality– relationship transactions. Thus, the tain relationship events, and certain relationship events occurrence of an event does not necessarily inform about may trigger developmental processes. To better under- the felt quality of the experience, and future research may stand these transactions, research is needed to (1) iden- test additional relationship characteristics to further ad- tify the broader, societal scripts, and expectations that are vance the understanding on transactions. associated with the relationship event, (2) determine the meaning that people ascribe to the relationship event, (3) account for the quality of the relationship experience, and 4.3.4 | Micro events and daily mechanisms (4) assess the daily experiences that are associated with the relationship event. While issues 1 to 3 emphasize that A final set that needs to be considered is the right tim- transactions may depend on more subtle, idiosyncratic ing to assess relationship events and personality traits (see conditions, issue 4 points to the necessity of considering also Luhmann et al.,  2014). Specifically, in the present the right timing between relationship events and person- work the time lag between personality assessments was ality assessments. Together, this knowledge will advance rather long and most of the assessed relationship events the understanding of the conditions, under which trans- occurred in the first (e.g., beginning a relationship) or last actions between traits and events occur across adulthood. (e.g., separation) years of the relationship (e.g., Denissen et al.,  2019). However, romantic relationships likely un- fold in the middle of these two endpoints, with the most 4.4 | Strengths and limitations crucial developmental period in the first ten years of the relationship (Bühler et al.,  2021). Therefore, research is In this work, we studied selection and socialization effects needed that more strongly considers the timing between between personality traits and relationship events with 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 16 | BÜHLER et al. representative, longitudinal data sets from three nations. of life– event research. A future path may be to ask in- The use of independent data sets as well as the analytic dividuals which event they think they have selected the approach of meta- analytic integration and internal repli- most (i.e., selection effects) and which event they think cation strengthens the validity of the conclusions (Duncan has shaped their personality the most (i.e., socialization et al., 2014; Orth et al., 2018). Still, some limitations need effects; Bleidorn et al., 2018). to be discussed. Fourth, similar to previous studies (see Table  S1), the First, although we assessed data sets that are repre- time lags between personality assessments were rather long sentative for people living in either Germany, Australia, in the present research, ranging from 4 to 6 years. Therefore, or the United Kingdom, the findings may not hold out- the present study generated knowledge about transactions side the Western culture. People across cultures may over longer time lags, but it could not inform about trans- differ in the timing, normativeness, and meaning of re- actions on a narrower time perspective. As noted, it would lationship events, which may have implications for the be useful to determine the immediate daily mechanisms, strength of transactions (Bleidorn et al., 2013; Gardiner by using experience sampling designs in the moment when et al., 2020). Moreover, some participants were clustered people transition into a new relationship event, such as in households (see Tables  S3−S5), which might have when partners are moving into a shared household (for an had some impact on selection and socialization effects. example study, see https://osf.io/u5sg3/). This fine- grained Specifically, regarding selection effects, given that fam- analysis of daily situations and experiences accompanying ily members share parts of genetic and environmental the relationship event might provide a more detailed un- variance, their personality might have been more similar derstanding of how transactions between personality traits (Kandler et al.,  2012) and so they might have selected and relationship events occur. Moreover, a higher num- more similar events. Regarding socialization effects, ber of assessments with a shorter density would allow to given that some people participated with their partner, also model nonlinear trajectories of personality develop- their relationship events might have been more similar ment, including temporary and sudden changes (Denissen (e.g., experiencing a messy, compared to an amicable, di- et al., 2019; Luhmann et al., 2014). vorce) and thus the effects of the events might have been more similar. However, we did not expect systematic ef- fects, given that only a small percentage of participants 5 | CONCLUSION were clustered in households. Second, in addition to studying selection and so- This research assessed transactions between personal- cialization effects, it might be worthwhile to examine ity traits and relationship events in three independent, anticipatory effects, that is, anticipatory personality de- nationally representative data sets, focusing on the role velopment before actually experiencing the relationship of age. The findings indicated that selection effects were event (Roberts et al.,  2004). Particularly in the case of more frequent, stronger, and more consistent than sociali- events that can be planned in advance (e.g., marriage), zation effects and that transactions with gain- based events social roles and psychological investments may happen were more prominent than transactions with loss- based before experiencing the event (Denissen et al.,  2019). events. Only few interactions with age were observed, However, given that anticipatory effects most likely and most of these interactions emerged with gain-b ased occur in the context of gain-b ased, rather than loss- events. Implications of the findings are that personality based, events (simply because loss-b ased events are often traits and relationship events are relatively weakly related harder to plan, at least not by both partners equally), we to each other across adulthood. This, however, may not did not examine anticipatory effects, but see promising necessarily mean that no effects are present, but that re- paths to study anticipatory effects from a developmental search is still in the process of identifying the most suitable perspective in future research. theoretical and methodological ways to comprehensively Third, in this research we set the focus on relationship understand personality–r elationship transactions. In this events as important markers of adulthood, consistent study, we discussed a series of influences that might mat- with previous research (e.g., Lodi-S mith & Roberts, 2007). ter for the strength of transactions and that might be ad- However, in many developmental periods relationship dressed in future research. events may coincide with occupational events, such as committing to both a long- term romantic relationship ACKNOWLEDGEMENTS and a serious job in young adulthood, while becoming The authors thank Louisa Scheling for her valuable help potentially widowed and retired in late adulthood. This, with coding as well as Laura Hansal and Konrad Hoppe in turn, makes it more difficult to unequivocally isolate for their support in preparing the tables. Open Access the effect of one life event, due to the naturalistic design funding enabled and organized by Projekt DEAL. 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BÜHLER et al. | 17 FUNDING INFORMATION REFERENCES This research was supported by Swiss National Science Allemand, M., Hill, P. L., & Lehmann, R. (2015). Divorce and personal- Foundation Grant P2BSP1_188102 to Janina Larissa ity development across middle adulthood. Personal Relationships, Bühler. Data, materials, and code are available on the 22(1), 122–1 37. https://doi.org/10.1111/pere.12067 Open Science Framework (OSF; https://osf.io/4cxhz/). Appelbaum, M., Cooper, H., Kline, R. B., Mayo- Wilson, E., Nezu, A. M., & & Rao, S. M. (2018). Journal article reporting stan- dards for quantitative research in psychology: The APA CONFLICT OF INTEREST Publications and Communications Board task force report. The authors have no conflict of interest to disclose. American Psychologist, 73(1), 3– 25. https://doi.org/10.1037/ amp00 00191 ETHICS STATEMENT Arnett, J. J. (2000). Emerging adulthood: A theory of development The nationally representative data sets used in this paper from the late teens through the twenties. American Psychologist, received ethical approval. 55(5), 469– 480. https://doi.org/10.1037/0003- 066x.55.5.469 Asendorpf, J. B., & Wilpers, S. (1998). Personality effects on social re- AUTHOR CONTRIBUTIONS lationships. Journal of Personality and Social Psychology, 74(6), 1531– 1544. https://doi.org/10.1037/0022- 3514.74.6.1531 Janina Larissa Bühler: conceptualization (lead), data Asselmann, E., & Specht, J. (2020). Taking the ups and downs at the management (lead), data analysis (lead), writing (lead). rollercoaster of love: Associations between major life events Marcus Mund: conceptualization (supporting), data anal- in the domain of romantic relationships and the Big Five per- ysis (supporting), writing (supporting). Franz J. Neyer: sonality traits. Developmental Psychology, 56(9), 1803– 1816. conceptualization (supporting), writing (supporting). https://doi.org/10.1037/dev000 1047 Cornelia Wrzus: conceptualization (supporting), writing Baltes, P. B. (1987). Theoretical propositions of life- span develop- (supporting). mental psychology: On the dynamics between growth and de- cline. Developmental Psychology, 23, 611– 626. Bleidorn, W., Hopwood, C. J., & Lucas, R. E. (2018). Life events and ORCID personality trait change. Journal of Personality, 86(1), 83– 96. Janina Larissa Bühler  https://orcid. https://doi.org/10.1111/jopy.12286 org/0000-0003-3684-9682 Bleidorn, W., Klimstra, T. A., Denissen, J. J. A., Rentfrow, P. J., Potter, Marcus Mund  https://orcid.org/0000-0003-0006-9043 J., & Gosling, S. D. (2013). Personality maturation around the Franz J. Neyer  https://orcid.org/0000-0001-8123-2011 world: A cross- cultural examination of social-i nvestment Cornelia Wrzus  https://orcid.org/0000-0002-6290-959X theory. Psychological Science, 24(12), 2530– 2540. https://doi. org/10.1177/09567 97613 498396 Bollen, K. A., & Curran, P. J. (2006). Latent curve models: A structural ENDNOTES equation approach. Wiley. 1 In this study, we did not include the gain- based event parenthood Borenstein, M., Higgins, J. P. T., Hedges, L. V., & Rothstein, H. R. because our main goal was to examine age effects on selection and (2017). Basics of meta-a nalysis: I2 is not an absolute measure of socialization effects. For that purpose, we had to ensure that all in- heterogeneity. Research Synthesis Methods, 8(1), 5–1 8. https:// cluded relationship events could theoretically occur across entire doi.org/10.1002/jrsm.1230 adulthood, which would not have been the case for parenthood Bühler, J. L., & Dunlop, W. L. (2019). The narrative identity approach (according to the World Health Organization  (2019), women's and romantic relationships. Social and Personality Psychology maximum reproductive age is 49 years). Compass, 13(4), e12447. https://doi.org/10.1111/spc3.12447 2 Given the focus of the present manuscript, this overview focuses Bühler, J. L., Krauss, S., & Orth, U. (2021). Development of rela- on transactions with Big Five personality traits, but transac- tionship satisfaction across the life span: A systematic review tions may also occur with surface characteristics of personality and meta- analysis. Psychological Bulletin, 147(10), 1012–1 053. (Kandler et al., 2014), such as with self- esteem and subjective https://doi.org/10.1037/bul00 00342 well- being (Luciano & Orth, 2017; Luhmann et al., 2012). Bühler, J. L., & Nikitin, J. (2020). Sociohistorical context and adult 3 For information on study protocol, data access, and publications social development: New directions for 21st century research. using this data set, see https://www.diw.de/en/soep.In the present American Psychologist, 75(4), 457–4 69. https://doi.org/10.1037/ study, we used Version 35 of the data set. amp000 0611 4 Caspi, A., & Moffitt, T. E. (1993). When do individual differences For information on study protocol, data access, and publica- matter? A paradoxical theory of personality coherence. tions using this data set, see https://melbo urnein stit ute.unime Psychological Inquiry, 4(4), 247– 271. https://doi.org/10.1207/ lb.edu.au/hilda. In the present study, we used Release 18 of the s15327 965pl i0404_1 data set. Caspi, A., Roberts, B. W., & Shiner, R. L. (2005). Personality 5 For information on study protocol, data access, and publications development: Stability and change. Annual Review of using this data set, see https://ww.unders tandi ngsoc iety.ac.uk/. In Psychology, 56(1), 453–4 84. https://doi.org/10.1146/annur the present study, we used the 13th Edition of the data set. ev.psych.55.090902.141913 6 A detailed description of the coding procedure is provided in the Census Bureau, U. S. (2021). Number, timing, and duration of mar- Supplemental Material (Part B) and on OSF. riages and divorces: 2016. Current Population Reports. https:// 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 18 | BÜHLER et al. www.census.gov/conte nt/dam/Censu s/libra ry/publi catio Campa (Eds.), Human bonding: The science of affectional ties ns/2021/demo/p70- 167.pdf (pp. 103– 131). The Guilford Press. Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of Haehner, P., Kritzler, S., Fassbender, I., & Luhmann, M. (2021). measurement invariance. Structural Equation Modeling, 14(3), Stability and change of perceived characteristics of major life 464– 504. https://doi.org/10.1080/107055 1070 1301834 events. Journal of Personality and Social Psychology, 122, 1098– Cheung, G. W., & Rensvold, R. B. (2002). valuating goodness- of- fit in- 1116. https://doi.org/10.1037/pspp0 000394 dexes for testing measurement invariance. Structural Equation Harris, K., & Vazire, S. (2016). On friendship development and the Modeling, 9(2), 233–2 55. https://doi.org/10.1207/S1532 8007S Big Five personality traits. Social and Personality Psychology EM0902_5 Compass, 10(11), 647– 667. https://doi.org/10.1111/spc3.12287 Costa, P. T., Jr., Herbst, J. H., McCrae, R. R., & Siegler, I. C. (2000). Havighurst, R. J. (1972). Developmental tasks and education (3rd ed.). Personality at midlife: Stability, intrinsic maturation, and re- McKay. sponse to life events. Assessment, 7(4), 365– 378. https://doi. Hofer, S. M., & Piccinin, A. M. (2009). Integrative data analysis org/10.1177/10731 91100 00700405 through coordination of measurement and analysis proto- Curran, P. J., & Hussong, A. M. (2009). Integrative data analysis: col across independent longitudinal studies. Psychological The simultaneous analysis of multiple data sets. Psychological Methods, 14(2), 150–1 64. https://doi.org/10.1037/a0015566 Methods, 14(2), 81– 100. https://doi.org/10.1037/a0015914 Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in co- Denissen, J. J. A., Luhmann, M., Chung, J. M., & Bleidorn, W. (2019). variance structure analysis: Conventional criteria versus new Transactions between life events and personality traits across alternatives. Structural Equation Modeling, 6, 1– 55. https://doi. the adult lifespan. Journal of Personality and Social Psychology, org/10.1080/107055 1990 9540118 116(4), 612– 633. https://doi.org/10.1037/pspp0 000196 Huedo- Medina, T. B., Sánchez- Meca, J., Marin- Martinez, F., & DeYoung, C. G. (2015). Cybernetic big five theory. Journal of Botella, J. (2006). Assessing heterogeneity in meta-a nalysis: Research in Personality, 56, 33– 58. https://doi.org/10.1016/j. Q statistic or I2 index? Psychological Methods, 11(2), 193– 206. jrp.2014.07.004 https://doi.org/10.1037/1082- 989X.11.2.193 Digman, J. M. (1997). Higher-o rder factors of the big five. Journal of Infurna, F. J., Gerstorf, D., & Lachman, M. E. (2020). Midlife in the Personality and Social Psychology, 73(6), 1246– 1256. https://doi. 2020s: Opportunities and challenges. American Psychologist, org/10.1037/0022- 3514.73.6.1246 75(4), 470– 485. https://doi.org/10.1037/amp00 00591 Duncan, G. J., Engel, M., Claessens, A., & Dowsett, C. J. (2014). Jackson, J. J., & Allemand, M. (2014). Moving personality develop- Replication and robustness in developmental research. ment research forward: Applications using structural equa- Developmental Psychology, 50(11), 2417– 2425. https://doi. tion models. European Journal of Personality, 28(3), 300– 310. org/10.1037/a0037996 https://doi.org/10.1002/per.1964 Dunlop, W. L., Hanley, G. E., McCoy, T. P., & Harake, N. (2017). John, O. P., Donahue, E. M., & Kentle, R. L. (1991). The “Big Five” Sticking to the (romantic) script: An examination of love life Inventory - Versions 4a and 54. Technical Report. University of scripts, stories, and self- reports of normality. Memory, 25(10), California, Institute of Personality and Social Research. 1444– 1454. https://doi.org/10.1080/09658 211.2017.1316509 John, O. P., & Srivastava, S. (1999). The Big Five trait taxonomy: Dyrenforth, P. S., Kashy, D. A., Donnellan, M. B., & Lucas, R. E. History, measurement, and theoretical perspectives. In L. A. (2010). Predicting relationship and life satisfaction from per- Pervin & O. P. John (Eds.), Handbook of personality: Theory and sonality in nationally representative samples from three coun- research (pp. 102– 138). Guilford Press. tries: The relative importance of actor, partner, and similarity Kandler, C., Bleidorn, W., Riemann, R., Angleitner, A., & Spinath, F. effects. Journal of Personality and Social Psychology, 99(4), 690– M. (2012). Life events as environmental states and genetic traits 702. https://doi.org/10.1037/a0020385 and the role of personality: A longitudinal twin study. Behavior Ebner, N. C., Freund, A. M., & Baltes, P. B. (2006). Developmental Genetics, 42(1), 57– 72. https://doi.org/10.1007/s1051 9- 011- 9491- 0 changes in personal goal orientation from young to late adult- Kandler, C., Zimmermann, J., & McAdams, D. P. (2014). Core and hood: From striving for gains to maintenance and prevention of surface characteristics for the description and theory of per- losses. Psychology and Aging, 21(4), 664–6 78. https://doi.org/10. sonality differences and development. European Journal of 1037/0882- 7974.21.4.664 Personality, 28, 231–2 43. https://doi.org/10.1002/per.1952 Gardiner, G., Baranski, E., & Bühler, J. L. (2020). Cross- cultural Karney, B. R., & Bradbury, T. N. (1995). Assessing longitudinal assessment of situational experience. In J. F. Rauthmann, R. change in marriage: An introduction to the analysis of growth A. Sherman, & D. C. Funder (Eds.), The Oxford handbook of curves. Journal of Marriage and the Family, 57(4), 1091– 1108. psychological situations. Oxford University Press. https://doi. https://doi.org/10.2307/353425 org/10.1093/oxfor dhb/978019 0263 348.013.11 Kazak, A. E. (2018). Editorial: Journal article reporting standards. Goebel, J., Grabka, M. M., Liebig, S., Kroh, M., Richter, D., American Psychologist, 73(1), 1– 2. https://doi.org/10.1037/ Schroeder, C., & Schupp, J. (2019). The German Socio- amp00 00263 Economic Panel (SOEP). Jahrbucher fur Nationalokonomie Kelly, E. L., & Conley, J. J. (1987). Personality and compatibility: A und Statistik, 239, 345– 360. https://doi.org/10.1515/jbnst prospective analysis of marital stability and marital satisfac- - 2018- 0022 tion. Journal of Personality and Social Psychology, 52(1), 27–4 0. Grimm, K. J., Ram, N., & Estabrook, R. (2016). Growth model- https://doi.org/10.1037/0022- 3514.52.1.27 ing: Structural equation and multilevel modeling approaches. Lehnart, J., & Neyer, F. J. (2006). Should I stay or should I go? Guilford Press. Attachment and personality in stable and instable romantic Günaydin, G., Selcuk, E., & Hazan, C. (2013). Finding the one: A relationships. European Journal of Personality, 20(6), 475–4 95. process model of human mate selection. In C. Hazan & M. I. https://doi.org/10.1002/per.606 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BÜHLER et al. | 19 Lehnart, J., Neyer, F. J., & Eccles, J. (2010). Long-t erm effects of Neugarten, B. L., Moore, J. W., & Lowe, J. C. (1965). Age norms, social investment: The case of partnering in young adult- age constraints, and adult socialization. American Journal of hood. Journal of Personality, 78(2), 639– 670. https://doi. Sociology, 7(6), 710–7 17. https://doi.org/10.1086/223965 org/10.1111/j.1467- 6494.2010.00629.x Neyer, F. J., & Asendorpf, J. B. (2001). Personality- relationship trans- Lewandowski, G. W., Jr., Aron, A., Bassis, S., & Kunak, J. (2006). action in young adulthood. Journal of Personality and Social Losing a self-e xpanding relationship: Implications for the self- Psychology, 81(6), 1190– 1204. https://doi.org/10.1037/002 concept. Personal Relationships, 13(3), 317–3 31. https://doi. 2- 3514.81.6.1190 org/10.1111/j.1475- 6811.2006.00120.x Neyer, F. J., & Lehnart, J. (2007). Relationships matter in personal- Lodi- Smith, J., & Roberts, B. W. (2007). Social investment and per- ity development: Evidence from an 8- year longitudinal study sonality: A meta- analysis of the relationship of personality across young adulthood. Journal of Personality, 75(3), 535– 568. traits to investment in work, family, religion, and volunteerism. https://doi.org/10.1111/j.1467- 6494.2007.00448.x Personality and Social Psychology Review, 11(1), 68– 86. https:// Neyer, F. J., Mund, M., Zimmermann, J., & Wrzus, C. (2014). doi.org/10.1177/108886 83062 94590 Personality- relationship transactions revisited. Journal Luciano, E. C., & Orth, U. (2017). Transitions in romantic relation- of Personality, 82(6), 539–5 50. https://doi.org/10.1111/ ships and development of self-e steem. Journal of Personality jopy.12063 and Social Psychology, 112(2), 307–3 28. https://doi.org/10.1037/ Oberski, D. (2014). lavaan.survey: An R Package for complex survey pspp00 00109 analysis of structural equation models. Journal of Statistical Luhmann, M., & Eid, M. (2009). Does it really feel the same? Changes Software, 57(1), 1– 27. https://doi.org/10.18637/j ss.v057.i01 in life satisfaction following repeated life events. Journal of OECD. (2019). Society at a glance 2019. OECD Publishing. Personality and Social Psychology, 97(2), 363– 381. https://doi. Orth, U., Erol, R. Y., & Luciano, E. C. (2018). Development of self- org/10.1037/a0015809 esteem from age 4 to 94 years: A meta-a nalysis of longitudinal Luhmann, M., Fassbender, I., Alcock, M., & Haehner, P. (2020). A studies. Psychological Bulletin, 144(10), 1045– 1080. https://doi. dimensional taxonomy of perceived characteristics of major org/10.1037/bul00 00161 life events. Journal of Personality and Social Psychology, 121(3), Pinquart, M., & Wahl, H. W. (2021). Subjective age from childhood to 633–6 68. https://doi.org/10.1037/pspp0 000291 advanced old age: A meta-a nalysis. Psychology and Aging, 36(3), Luhmann, M., Hofmann, W., Eid, M., & Lucas, R. E. (2012). 394–4 06. https://doi.org/10.1037/pag00 00600 Subjective well- being and adaptation to life events: A meta- Pusch, S., Mund, M., Hagemeyer, B., & Finn, C. (2019). Personality analysis. Journal of Personality and Social Psychology, 102(3), development in emerging and young adulthood: A study of 592– 615. https://doi.org/10.1037/a0025948 age differences. European Journal of Personality, 33, 245–2 63. Luhmann, M., Orth, U., Specht, J., Kandler, C., & Lucas, R. E. (2014). https://doi.org/10.1002/per.2181 Studying changes in life circumstances and personality: It's R Development Core Team. (2020). R: A language and environ- about time. European Journal of Personality, 28(3), 256– 266. ment for statistical computing. R Foundation for Statistical Magnusson, D. (1990). Personality development from an interac- Computing. https://www.R-p rojec t.org tional perspective. In L. A. Pervin (Ed.), Handbook of person- Roberts, B. W., & Bogg, T. (2004). A longitudinal study of the ality: Theory and measurement (pp. 193– 222). Guilford Press. relationships between conscientiousness and the social- Mayer, K. U. (2009). New directions in life course research. Annual environmental factors and substance- use behaviors that influ- Review of Sociology, 35, 413– 433. https://doi.org/10.1146/annur ence health. Journal of Personality, 72(2), 325– 354. https://doi. ev.soc.34.040507.134619 org/10.1111/j.0022-3 506.2004.00264.x McCrae, R. R., & Costa, P. T., Jr. (1997). Conceptions and correlates Roberts, B. W., O'Donnell, M., & Robins, R. W. (2004). Goal and per- of openness to experience. In Robert Hogan, Stephen Briggs, sonality trait development in emerging adulthood. Journal of John Johnson (Eds.), Handbook of personality psychology (pp. Personality and Social Psychology, 87(4), 541– 550. https://doi.or 825– 847). Academic Press. https://doi.org/10.1016/B978- 01213 g/10.1037/0022- 3514.87.4.541 4645- 4/50032- 9 Roberts, B. W., Walton, K. E., & Viechtbauer, W. (2006). Patterns of McCrae, R. R., & Costa, P. T., Jr. (2008). The five-f actor theory of mean- level change in personality traits across the life course: personality. In O. P. John, R. W. Robins, & L. A. Pervin (Eds.), A meta- analysis of longitudinal studies. Psychological Bulletin, Handbook of personality: Theory and research (pp. 159–1 81). 132(1), 1–2 5. https://doi.org/10.1037/0033-2 909.132.1.1 Guilford Press. Roberts, B. W., & Wood, D. (2006). Personality development in the Melbourne Institute of Applied Economic and Social Research. context of the neo-s ocioanalytic model of personality. In D. K. (2017). The Household, Income and Labour Dynamics in Australia Mroczek & T. D. Little (Eds.), Handbook of personality develop- (HILDA) Survey: General release 16 (Waves 1–1 6). Department of ment (pp. 11– 39). Erlbaum. Social Services, Melbourne Institute of Applied Economic and Roberts, B. W., Wood, D., & Caspi, A. (2008). The development of Social Research. https://doi.org/10.4225/87/VHRTR5 personality traits in adulthood. In O. P. John, R. W. Robins, & L. Mund, M., Jeronimus, B. F., & Neyer, F. (2018). Personality and so- A. Pervin (Eds.), Handbook of personality: Theory and research cial relationships: As thick as thieves. In C. Johansen (Ed.), (pp. 375– 398). Guilford Press. Your personality makes you ill: Scientific proof or wishful think- Roberts, B. W., Wood, D., & Smith, J. L. (2005). Evaluating five factor ing? (pp. 153– 183). Elsevier. theory and social investment perspectives on personality trait Mund, M., Lüdtke, O., & Neyer, F. (2020). Owner of a lonely heart: development. Journal of Research in Personality, 39(1), 166– The stability of loneliness across the life span. Journal of 184. https://doi.org/10.1016/j.jrp.2004.08.002 Personality and Social Psychology, 119(2), 497– 516. https://doi. Rosseel, Y. (2012). Lavaan: An R package for structural equation org/10.1037/pspp00 00262 modeling and more. Journal of Statistical Software, 48, 1–3 6. 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 20 | BÜHLER et al. Rubin, D. C., Berntsen, D., & Hutson, M. (2009). The normative and Wagner, G. G., Frick, J. R., & Schupp, J. (2007). The German the personal life: Individual differences in life scripts and life Socioeconomic Panel Study (SOEP): Scope, evolution and en- story events among USA and Danish undergraduates. Memory, hancements. Journal of Applied Social Science Studies, 127, 17(1), 54– 68. https://doi.org/10.1080/096582 10802 541442 139–1 69. Saucier, G. (1994). Mini- markers: A brief version of Goldberg's uni- Wagner, J., Becker, M., Ludtke, O., & Trautwein, U. (2015). The polar big-f ive markers. Journal of Personality Assessment, 63(3), first partnership experience and personality development: A 506– 516. https://doi.org/10.1207/s15327 752j pa6303_8 propensity score matching study in young adulthood. Social Scheling, L., & Richter, D. (2021). Generation Y: Do millennials Psychological and Personality Science, 6(4), 455– 463. https:// need a partner to be happy? Journal of Adolescence, 90, 23–3 1. doi.org/10.1177/19485 506145 66092 https://doi.org/10.1016/j.adole scence.2021.05.006 Widaman, K. F., Ferrer, E., & Conger, R. D. (2010). Factorial invari- Schupp, J., & Gerlitz, J.- Y. (2014). Big Five Inventory- SOEP (BFI-S ). ance within longitudinal structural equation models: Measuring Zusammenstellung sozialwissenschaftlicher Items und Skalen the same construct across time. Child Development Perspectives, (ZIS). https://doi.org/10.6102/zis54 4(1), 10– 18. https://doi.org/10.1111/j.1750- 8606.2009.00110.x Smilie, L. D. (2013). Extraversion and reward processing. Current Wilt, J., & Revelle, W. (2017). Extraversion. In T. A. Widiger (Ed.), Directions in Psychological Science, 22(3), 167–1 72. https://doi. The Oxford handbook of the five factor model (pp. 57–8 1). org/10.1177/09637 214124 70133 Oxford University Press. Solomon, B. C., & Jackson, J. J. (2014). Why do personality traits pre- World Health Organization. (2019). Maternal, newborn, child and dict divorce? Multiple pathways through satisfaction. Journal of adolescent health and aging. World Population Prospects. Personality and Social Psychology, 106(6), 978–9 96. https://doi. Wrzus, C., Luong, G., Wagner, G. G., & Riediger, M. (2021). org/10.1037/a0036190 Longitudinal coupling of momentary stress reactivity and trait Specht, J., Egloff, B., & Schmukle, S. C. (2011). Stability and change neuroticism: Specificity of states, traits, and age period. Journal of personality across the life course: The impact of age and of Personality and Social Psychology, 121(3), 691– 706. https:// major life events on mean-l evel and rank- order stability of the doi.org/10.1037/pspp0 000308 Big Five. Journal of Personality and Social Psychology, 101(4), Wrzus, C., & Neyer, F. J. (2016). Co- development of personality and 862– 882. https://doi.org/10.1037/a0024950 friendships across the lifespan: An empirical review on selec- Spikic, S., Mortelmans, D., & Pasteels, I. (2021). Does divorce change tion and socialization. European Psychologist, 21(4), 254– 273. your personality? Examining the effect of divorce occurrence https://doi.org/10.1027/1016- 9040/a000277 on the Big Five personality traits using panel surveys from three Wrzus, C., & Roberts, B. W. (2017). Processes of personality devel- countries. Personality and Individual Differences, 171, 110428. opment in adulthood: The TESSERA framework. Personality https://doi.org/10.1016/j.paid.2020.110428 and Social Psychology Review, 21(4), 253– 277. https://doi. Statista. (2021). Durchschnittliches Heiratsalter von Männer und org/10.1177/108886 83166 52279 Frauen in Deutschland von 1991 bis 2020. https://de.stati sta. com/stati stik/daten/ studi e/18032 1/umfra ge/durch schni ttlic hes-a lter - bei- der- ehesch lies sung-n ach- geschl echt/ SUPPORTING INFORMATION Stephan, Y., Sutin, A. R., & Terracciano, A. (2014). Subjective age Additional supporting information can be found online and personality development: A 10- Year study. Journal of Personality, 83(2), 142– 154. https://doi.org/10.1111/jopy.12090 in the Supporting Information section at the end of this van Scheppingen, M. A., Jackson, J. J., Specht, J., Hutteman, R., article. Denissen, J. J. A., & Bleidorn, W. (2016). Personality trait de- velopment during the transition to parenthood: A test of social investment theory. Social Psychological and Personality Science, How to cite this article: Bühler, J. L., Mund, M., 7(5), 452– 462. https://doi.org/10.1177/194855 0616 630032 Neyer, F. J., & Wrzus, C. (2022). A developmental Viechtbauer, W. (2005). Bias and efficiency of meta-a nalytic vari- perspective on personality–r elationship ance estimators in the random- effects model. Journal of transactions: Evidence from three nationally Educational and Behavioral Statistics, 30(3), 261–2 93. https:// representative samples. Journal of Personality, 00, doi.org/10.3102/10769 98603 0003261 1–20. https://doi.org/10.1111/jopy.12757 Viechtbauer, W. (2010). Conducting meta-a nalyses in R with the metafor package. Journal of Statistical Software, 36(3), 1– 48. https://doi.org/10.18637/ jss.v036.i03 14676494, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jopy.12757 by Universitätsbibliothek Mainz, Wiley Online Library on [27/01/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License